stock-analysis
Pass
Audited by Gen Agent Trust Hub on Oct 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill fetches and processes external news headlines from Google News RSS to detect geopolitical and economic risks.
- Ingestion points: The
check_breaking_newsfunction inscripts/analyze_stock.pyretrieves content from Google News RSS feeds. - Boundary markers: External headline data is interpolated into the output report without specific protective delimiters.
- Capability inventory: The skill can write to a local portfolio JSON file and performs network GET requests to fetch financial data.
- Sanitization: Headlines are lowercased and matched against a hardcoded list of
CRISIS_KEYWORDS. The text is used for numerical scoring and informative caveats rather than being interpreted by the agent as direct instructions, which significantly limits exploitability. - [EXTERNAL_DOWNLOADS]: Fetches financial data, sentiment indicators, and news from reputable and well-known services including Yahoo Finance, CNN Business, SEC EDGAR, and Google News.
- [COMMAND_EXECUTION]: The skill uses the
uvtool to execute local Python scripts for stock analysis and portfolio management as its core functionality.
Audit Metadata